ABSTRACT Functions of many enzymes and signaling proteins strongly depend on allostery, a phenomenon of dynamical coupling between distant sites of a biological macromolecule. Allosterically regulated proteins are usually modulated by either ligand binding or post-translational modifications. Understanding the structural and dynamic underpinnings of allosteric regulation and, especially, learning how to engineer allostery provides an outstanding opportunity to control protein function and cellular activity. The ability to control cell phenotype through engineered allosteric regulation of guanine exchange factors (GEFs), proto-oncogene tyrosine-protein kinase (Src), guanosine triphosphate hydrolases (GTPases) has great potential in designing therapies, developing biosensors/biomarkers for early disease diagnosis, as well as providing a general tool for fundamental biological research. While remarkable progress has been achieved in understanding how allostery between distant residues is maintained from theoretical, experimental and evolutionary perspectives, we still lack a clear picture of the inter-residue interaction networks in proteins. Our goal is to develop and experimentally validate an efficient computational methodology that enables the identification and alteration of allosteric interaction networks in proteins. We hypothesize that protein activity can be controlled through the modulation of the dynamics of distant but allosterically coupled regions by exploiting the interaction networks among protein residues. To test our hypothesis, we propose to develop a general and efficient computational methodology for the engineering of allosteric regulation based on molecular dynamics simulations and graph-theoretical approaches (aim 1). We will tune and validate our computational analysis using proteins whose allostery has been experimentally determined. Also we will predict allosteric networks for a series of test proteins and after will use NMR methods to verify our computation predictions. Further, we propose immediate applications that will benefit several fields of research. We will insert molecular switches featuring either light- or drug-induced conformational changes into identified allosteric sites, and utilize the allosteric coupling network to control the conformational dynamics of the active site, either inhibiting or activating protein function. Specifically, we will design a photo-inhibited small GTPase, CDC42 (aim 2), the activity of which will be reversibly controlled with pulses of blue light in living cells. For this construct, CDC42 plus the switch, allosteric pathways and intrinsic dynamic features will be mapped by NMR methods that will further aid the computational strategies employed for manipulating allostery. We will employ this designed GTPases to study the role of CDC42 in cell motility, thereby proving our ability to design allosteric regulators, and producing a practical and valuable tool for cell biology and kinetics studies. Next, we will design an allosterically regulated analog of the guanine exchange factor Vav2 and Src kinase (aim 3), which are critical to multiple signaling pathways, and involved in homeostasis, cancer and developmental diseases. This aim will demonstrate transferability of our methodology to big protein families as GTPases, GEFs and kinases.